Abstract
Previous research has investigated the relationship between cultural values and leadership. This research expands on this tradition and examines how the strength of social norms—or tightness–looseness—influences perceptions of effective leadership. Data from Gelfand, Raver, et al. were integrated with GLOBE’s leadership research to examine the attributes of leaders seen as leading to effectiveness in tight and loose cultures. Analyses of data across 29 samples show that cultural tightness is positively related to the endorsement of autonomous leadership and negatively related to the endorsement of charismatic and team leadership, even controlling for in-group collectivism, power distance, and future orientation at the societal and organizational level of analysis. Theoretical and practical implications are discussed.
Leadership, the process by which one person or a group of individuals influences others, has been of interest to social scientists and policy makers alike (Bass, 1990; House & Aditya, 1997). Many theoretical frameworks have sought to understand the complex interrelationships between leaders and followers (House & Aditya, 1997; Jago, 1982). Historically, much of the research on these models was developed and tested solely in Western contexts. More recently, leadership research has gone global and has begun to illuminate what is etic (universal) and emic (culture specific) regarding the leadership attributes that are perceived as effective versus ineffective (Aktaş & Sargut, 2011; Dickson, Den Hartog, & Mitchelson, 2003; Dorfman & House, 2004; House, Hanges, Javidan, Dorfman, & Gupta, 2004; House, Wright, & Aditya, 1997; Sargut & Aktaş, 2012). Taking an open systems perspective (Katz & Kahn, 1970), this line of research has shown that leadership processes do not occur in a vacuum; rather, they are affected by the broader organizational and societal cultural values in which organizations are embedded (for reviews, see Dickson et al., 2003; House et al., 2004).
Although leadership research is now being expanded beyond Western borders, it is important to note that the field has inadvertently focused almost exclusively on one aspect of culture—cultural values such as individualism-collectivism—while neglecting other aspects of culture, leading scholars to urge that we expand our “conceptual toolkit” to broaden the scope of cross-cultural leadership research (Bond, 1997; Earley & Mosakowski, 2002; Gabrenya, 1999; Gelfand, Nishii, & Raver, 2006; Ip & Bond, 1995). In the current article, we begin to fill this void and complement existing research by examining the influence of the strength of social norms, or cultural tightness–looseness (Gelfand et al., 2006; Gelfand et al., 2011; Harrington & Gelfand, 2014; Pelto, 1968; Triandis, 1989), on perceptions of effective leadership.
Cultural tightness–looseness has its theoretical roots in multiple disciplines, including anthropology (Pelto, 1968), sociology (Boldt, 1978a, 1978b), and psychology (Berry, 1966, 1967), and contrasts cultures that have strong norms and little tolerance for deviance with those that have weak norms and high tolerance for deviance (Gelfand et al., 2006; Gelfand et al., 2011; Harrington & Gelfand, 2014; Roos, Gelfand, Nau, & Lun, 2015; Triandis, 1989). Research has shown that nations vary widely in tightness–looseness and that the construct is distinct from cultural values (Carpenter, 2000; Gelfand et al., 2011). Tightness–looseness has been linked to numerous organizational outcomes in a recent meta-analysis (Taras, Kirkman, & Steel, 2010), as well as CEO discretion (Crossland & Hambrick, 2011), organizational creativity (Chua, Roth, & Lemoine, 2015; Ozeren, Ozmen, & Appolloni, 2013), negotiation (Gunia, Brett, Nandkeolyar, & Kamdar, 2011), expatriates’ job satisfaction (Peltokorpi & Froese, 2014), and even stock price synchronicity (Eun, Wang, & Xiao, 2015) and terrorism (Gelfand, LaFree, Fahey, & Feinberg, 2013).
Despite the link between tightness–looseness and numerous organizational outcomes, to our knowledge, this is the first research to examine the influence of tightness–looseness on perceptions of effective leadership (see Toh & Leonardelli, 2012, who examined the impact of tightness–looseness on the emergence of female leaders). Our research, described below, illustrates that tightness–looseness has distinct implications for the type of leadership that is perceived to be effective across cultures, and accounts for unique variance in perceived effectiveness of leadership attributes, above and beyond other cultural dimensions.
In what follows, we first discuss previous research on cultural tightness–looseness. We then put forth new theory and hypotheses—informed by the cognitive tradition in leadership research—relating tightness–looseness to implicit theories of autocratic, charismatic, self-protective, autonomous, team, and humane-oriented leadership. Using multilevel modeling, we test these relationships across 29 samples while controlling for other cultural dimensions at the societal and organizational levels of analysis. We conclude with theoretical and practical implications of this research.
Tightness–Looseness as a Fundamental Dimension of Cultural Variation
In his classic article, “The Difference Between ‘Tight’ and ‘Loose’ Societies,” Pelto (1968) showed that traditional societies varied widely in their expression of and adherence to social norms. “Tight” societies were described as those that were rigorously formal and disciplined, had clearly defined norms, and imposed severe sanctions on individuals who deviate from norms. “Loose” societies were described as those that had a lack of formality, regimentation and discipline, had norms expressed through a wide variety of alternative channels, and had a high tolerance for deviant behavior (Pelto, 1968). The Hutterites, Hanno communities, Lubara, and the Israeli Kibbutz were ranked as among the tightest societies, with very strong norms and little tolerance for deviant behavior, whereas the Kung Bushman, Cubeo, and the Skolt Lapps were rated as the among loosest, with ambiguous norms and a high tolerance for deviant behavior (Pelto, 1968). Later anthropological work showed that agricultural societies were tighter than hunting and gathering cultures, given that the former require norms to foster the coordination necessary to grow crops for survival (Lomax & Berkowitz, 1972; see also Barry, Child, & Bacon, 1959, McNett, 1970). In all, this line of anthropological work not only described cultural variation in tightness–looseness, but also began to identify particular antecedents of this cultural dimension in traditional cultures (see also Boldt, 1978a, 1978b; Boldt & Roberts, 1979).
Within psychology, Berry and his colleagues (Berry, 1966, 1967; Witkin & Berry, 1975) were among the first to identify psychological correlates of tightness–looseness. Berry (1966) showed that individuals in tightly structured agricultural settings (e.g., the Temne of Sierra Leone) exhibited lower psychological differentiation (i.e., a reduced sense of separation of the self from others (Witkin, 1949; Witkin & Berry, 1975) as compared with individuals in loosely structured hunting and fishing settings (e.g., Eskimos). Dawson (1967a, 1967b) likewise found that in groups that had strict discipline (e.g., the Temne), children were more likely to develop a field-dependent cognitive style, whereas groups that had more lenient childrearing practices (e.g., the Mende of Sierra Leone) were more likely to develop a field independent cognitive style. After a long hiatus of research on the topic, Triandis (1989) reintroduced the tightness–looseness construct in a seminal article in Psychological Review, wherein he clearly differentiated the construct of tightness from collectivism. This prediction was indeed supported later in a study of traditional cultures by Carpenter (2000). Consistent with Triandis’s (1989) premonition, Carpenter (2000) found that that constructs were only moderately correlated (r = .44).
Although discussions related to tightness–looseness date back to the 1960s, and the dimension has received some attention in anthropology, sociology, and psychology, it is only recently that there has been systematic study of tightness–looseness in modern cultures. In their multilevel study across 33 nations, Gelfand and colleagues (2011) showed that nations vary widely on tightness–looseness. Situating their work within a larger eco-cultural framework (Triandis, 1972; Witkin & Berry, 1975), they found that tight cultures have greater population density, a dearth of natural resources (e.g., more food deprivation, less food supply, less access to safe water and air, and lower percentage of farmland), greater environmental threats (e.g., natural disaster vulnerability), and greater health vulnerabilities (e.g., higher life lost due to communicable or contagious diseases and higher infant mortality). More generally, the authors identified clear norms and strict norm enforcement as adaptive mechanisms for survival and social coordination, which can help cultures to cope with environmental pressure, resource scarcity, and disease containment (see Roos et al., 2015, for evolutionary game theoretic models supporting this notion). Tightness–looseness is also related to the strength of socio-political institutions (e.g., autocracy, openness of the media) and constraints in everyday situations. At the individual level, tightness is related to higher felt accountability; individuals in tight cultures generally have more of a prevention focus, higher self-control, and greater self-monitoring, whereas individuals in loose cultures have more of a promotion focus, and less self-control and self-monitoring ability (Gelfand et al., 2006), attributes which help individuals to fit in and maintain the strength of situations in their cultural context. 1 In a recent study, Harrington and Gelfand (2014) showed that many of the same correlates of tightness hold within-nations at the state level of analysis.
As noted above, in recent years, tightness–looseness has been related to a wide range of organizational phenomena, but has yet to be examined for its effects on perceptions of effective leadership. Drawing on implicit leadership theory (Den Hartog, House, Hanges, Ruiz-Quintanilla, & Dorfman, 1999; Lord, Foti, & De Vader, 1984; Lord, Foti, & Phillips, 1982; Phillips & Lord, 1981), we next set forth predictions regarding the leadership attributes that are seen as most effective in tight and loose cultures.
Culture and Implicit Leadership Theory
Implicit leadership draws on research in cognitive psychology and posits that leadership is understood by a cognitive categorization process by which individuals categorize the attributes that distinguish between effective and ineffective leaders (Den Hartog et al., 1999; Lord, et al., 1984; Lord et al., 1982; Phillips & Lord, 1981). According to this approach, individuals match the perceived attributes of leaders with their internal categories of leadership attributes to judge whether they are effective (Foti & Luch, 1992). Lord and colleagues, in particular, proposed three levels of leadership categorization. The highest and most general category is the “superordinate level,” in which people categorize attributes common to most leaders (Lord et al., 1984; Lord et al., 1982). In the second category, people refine leadership in terms of situational and contextual characteristics such as religious, political, or business leaders. And in the basic subordinate level, types of leaders within a context are differentiated such as left- versus right-wing political leaders or upper versus middle-level managers.
Most pertinent to this research, culture has important effects on the development of superordinate categories of effective leadership prototypes, or people’s categorization of the attributes that are common to leaders (Hunt, Boal, & Sorenson, 1990; Lord & Maher, 1991). Gerstner and Day (1994) were among the first to show evidence of cultural differences in the perception of leadership prototypes. In their study, they investigated variance in the prototypicality of 59 leadership attributes among students from seven countries and found that leadership prototypes varied by the respondents’ country of origin. Later, the GLOBE Global Leadership and Organizational Behavior Effectiveness (GLOBE) research team (House et al., 2004) conducted a large scale study on how the leadership prototypes differ according to organizational and national cultural values across 62 societies. They introduced six superordinate dimensions of leadership that are made up from 21 second-order leadership dimensions (Hanges & Dickson, 2004; House et al., 2004). The first of these leadership dimensions is charismatic leadership, which includes visionary, inspirational, self-sacrifice, integrity, decisive, and performance-oriented primary leadership dimensions (Hanges & Dickson, 2004). The second dimension is team-oriented leadership, which includes the primary leadership dimensions of collaborative, team orientation, team integrator, diplomatic, malevolent (reverse scored), and administratively competent (Hanges & Dickson, 2004). The third dimension is self-protected leadership, which is composed of self-centered, status consciousness, conflict inducer, face-saver, and procedural leadership dimensions (Hanges & Dickson, 2004). The fourth dimension is participative leadership, which is composed of autocratic and participative leadership primary leadership dimensions (Hanges & Dickson, 2004). The fifth dimension is humane-oriented leadership, which is made up of modesty and humane-orientation primary leadership dimensions (Hanges & Dickson, 2004). The final dimension is autonomous leadership, which focuses on independent leadership and not relying on others to make decisions (Hanges & Dickson, 2004).
In this article, we integrate previous research on tightness–looseness with the GLOBE’s leadership framework to examine, for the first time, the connection between tightness–looseness and perceptions of effective leadership. In order to illustrate whether tightness–looseness adds to our understanding of cross-cultural leadership beyond existing value dimensions, we take a conservative approach and control for in-group collectivism and future orientation at the society level given that these values have been shown to have moderately strong relationships with tightness–looseness (Gelfand et al., 2011). We also control for power distance given it is critical to explain superior subordinate relations in a society (Hofstede, 1980;1991). Controlling for these cultural dimensions at the societal level enables us to investigate whether cultural tightness has a unique contribution to the understanding of perceptions of leadership effectiveness over and beyond these existing societal dimensions. We also control for these dimensions at the organizational level, as this allows us to investigate the unique contribution of the national level variables more accurately as leaders also adapt their leadership styles according to organizational culture (Schein, 1992), and individuals’ expectations for leadership also vary by organizational culture (House et al., 2004).
Hypotheses: Cultural Tightness–Looseness and Perceptions of Effective Leadership
We first explore the relationship between cultural tightness and participative leadership. Participative leadership is a dimension that reflects the extent to which the managers involve others in making and implementing decisions (Dorfman, Hanges, & Brodbeck, 2004). Participative leaders give others the opportunity to fulfill their own needs and to self-actualize (Jago, 1982). By contrast, autocratic leadership is characterized by highly centralized decision making and concentrated authority (Jago, 1982). Autocratic leaders structure tasks for subordinates and clarify expectations that need to be followed (Rast, Gaffney, Hogg, & Crisp, 2012). In doing so, they restrict the range of permissible behavior. Cultural tightness is related to the expectation that there are clear rules and little tolerance for deviance and to the need for structure, which theoretically forms a basis for the emergence of autocratic leadership (Kruglanski, Pierro, Mannetti, & De Grada, 2006; Pierro, Cicero, Bonaiuto, van Knippenberg, & Kruglanski, 2005). Hence, it can be theorized that participative leadership is more congruent with practices and conditions in loose compared with tight cultures. Indirect support for this notion can be found in Gelfand et al.’s (2011) analysis of the World Value Survey. These authors found that individuals in tighter countries think that the responsibility of government is to maintain order and viewed army-controlled governments more favorably as compared with individuals in loose cultures:
A new aspect of leadership, identified by GLOBE, is autonomous leadership. Autonomous leaders are defined as leaders who have extreme confidence in their own abilities and less confidence in the abilities and ideas of others (House et al., 2014; House et al., 2004). Because of this belief system, such leaders tend to work independently, without collaboration or feedback from others (House et al., 2014). Prior research has found that these leadership characteristics are believed to be more effective in societies that have more assertive and uncertainty avoidant cultural practices (Dorfman et al., 2004). As a result of these leaders being disconnected from their advisors, as well as from organizational feedback provided by and about others, these leaders are theorized to be less innovative and reinforce the status quo as compared with other types of leaders. Thus, we predict that such leadership would be desired in culturally tight societies because such societies also tend to resist changes in the status quo. Having strong confidence in one’s abilities would also presumably be preferred in contexts where there is high threat, which often characterizes tight cultures:
We next examine the connection between tightness–looseness and charismatic leadership. Charismatic leaders demand a high standard of performance, use innovative means to achieve their goals, and engage in revolutionary, transformational, and visionary behaviors to motivate followers (Conger & Kanungo, 1987; Dorfman et al., 2004). Tight and loose cultures differ in their preference for stability versus change and adaptation versus innovation (Chua et al., 2015; Gelfand et al., 2006). In tight cultures, individuals are generally more prevention focused and are reluctant to change the status quo (Gelfand et al., 2006). Individuals in loose cultures, by contrast, generally prefer to come up with new ideas that challenge the status quo (Chua et al., 2015; Gelfand et al., 2006). Given that tightness is related to preserving status quo, whereas looseness is related to innovation, we predicted that people in loose cultures would perceive charismatic leaders as more effective than people in tight cultures:
Self-protective leadership focuses on safety and security of the individual leader or group (Dorfman et al., 2004). Such leaders have a desire to succeed among possible competitors for the leader’s position and success (House et al., 2014). These leaders tend to engage in ritualistic, formal, cautious, and habitual behaviors (Brodbeck, Frese, & Javidan, 2002). Gelfand et al. (2006) proposed that in tight cultures, the preferred decision-making style is an adaptive style, which is characterized as cautious, reliable, efficient, and disciplined. In contrast, in loose cultures, the preferred mode of decision making is theorized to be an innovative style (Gelfand et al., 2006). Individuals who have an innovative thinking style prefer to challenge established rules and procedures, ignore the constraints of prevailing paradigms, and derive their ideas for solutions from outside of the system (Gelfand et al., 2006; Kirton, 1976). Hence, we expected that the procedural, cautious, ritualistic, and risk avoidance nature of self-protected leadership (Dorfman et al., 2004) will be perceived as more effective in tight than in loose cultures:
In addition to these dimensions, we also examined the relationship between tightness–looseness and the two remaining GLOBE leadership dimensions: team orientation and humane orientation. Team-oriented leaders are loyal to and care for the welfare of their team members (House et al., 2014). They use their skills to manage the internal dynamics of their team to build a more cohesive group (House et al., 2014). Given the close connection of team-oriented leadership to group orientation, we expected it to be more highly related to collectivism than tightness. Finally, humane leaders are described as unpretentious and reticent to boast. They are empathetic toward others and very likely to offer help and assistance to others. Humane-oriented leadership represents a person’s emphasis on building relationship with others, which is theoretically more related to relationship orientation and collectivism.
Method
Sample
The data for this study were created by combining country-level tightness scores from Gelfand et al. (2011) with GLOBE’s individual, organizational, and societal level data that measure leadership preferences, organizational culture, and societal culture (House et al., 2004). The Gelfand et al.’s (2011) tightness data consisted of 6,823 participants from 33 nations. Approximately 58.6% of the sample was female. The mean age of participants was 30.1 years, and the average amount of work experience was 8 years (Gelfand et al., 2011). In the 2004 GLOBE study, data were collected from 15,247 middle-level managers nested in 759 organizations from the financial, food processing, and telecommunications industries in 62 societies. The number of respondents per country ranged from 27 to 1,790, with an average of 250 per country (House & Hanges, 2004). The matching countries and sample size by country for the tightness and Globe studies are shown in Table 1. 2
Sample Size by Country and Study.
Measures
Cultural tightness was measured using the six-item scale developed by Gelfand et al. (2011). The items on this scale assessed the clarity and number of social norms, the degree of tolerance for norm violations, and overall compliance with social norms in each nation. Sample items include “There are many social norms that people are supposed to abide by in this country”; “In this country, if someone acts in an inappropriate way, others will strongly disapprove”; and “People in this country almost always comply with social norms.” Procrustes analyses demonstrated that the scale had metric equivalence across cultures. The reliability of the scale was .85, and people in tight and loose cultures also agree on the levels of tightness–looseness in their respective nations as indicated by the mean Rwg(j)of 0.85, the ICC(1) of .13 and the ICC(2) of .97. Details on the scale can be found in Gelfand et al. (2011).
Data regarding societal and organizational culture on collectivism, power distance, and future orientation were obtained from the GLOBE study (House et al., 2004). Specifically, the GLOBE practices scales (i.e., the country “as is”) were used given that they are more highly related to tightness–looseness than GLOBE “should be” scales. Separate organizational and societal scales were created for each cultural dimension (Hanges & Dickson, 2004). Sample items for the societal scales include “In this society, children take pride in the accomplishments of their parents (in-group collectivism)”; “In this society, rank and position in the hierarchy have special privileges (power distance)”; and “The way to be successful in this society is to plan ahead (1) or take life as they occur (7) (future orientation).” Sample items for the organizational scales include “In this organization, group members take pride in the individual accomplishments of their group manager (in-group collectivism)”; “In this organization, a person’s influence is based primarily on one’s ability and contribution to the organization (1) or the authority of one’s position (7) (power distance)”; and “In this organization, the accepted norm is to plan for the future (1) or to accept the status quo (7) (future orientation).” (For more detailed information about the GLOBE scales and the methodology used to develop them, see Hanges & Dickson, 2004; House & Hanges, 2004.) As discussed by Hanges and Dickson (2004), support was found for the aggregation of the culture scales to their desired level of analysis as well as for the unidimensionality of the items within each scale. The internal consistencies for the societal and organizational cultural practices scales were generally acceptable (i.e., Cronbach’s αIn-Group Collectivism–Societal = .77; Cronbach’s αIn-Group Collectivism–Organizational = .70; Cronbach’s αFuture Orientation–Societal = .80; Cronbach’s αFuture Orientation–Organizational = .67; Cronbach’s αPower Distance–Societal = .80; Cronbach’s αPower Distance–Organizational = .57). The reliability of the culture means also exhibited substantial consistency (i.e., ICC(2)In-Group Collectivism–Societal = .98; ICC(2)In-Group Collectivism–Organizational = .91; ICC(2)Future Orientation–Societal = .92; ICC(2)Future Orientation–Organizational = .91; ICC(2)Power Distance–Societal = .91; ICC(2)Power Distance–Organizational = .92). 3
Finally, data regarding the content of shared leadership schemas were obtained using the data from GLOBE culturally endorsed leadership theory (CLT) scales. Shared leadership schemas were measured by using 112 leadership attribute items. Respondents were asked to rate leadership behaviors on a scale of 1 = this behavior or characteristic greatly inhibits a person from being an outstanding leader to 7 = this behavior or characteristic contributes greatly to a person being an outstanding leader. The 112 items assessed 21 primary leadership dimensions (e.g., inspirational, diplomatic, modesty), which were subsequently combined into six second-order leadership dimensions, or what GLOBE called global leadership dimensions (Hanges & Dickson, 2004). Sample items include “Makes decisions in a dictatorial way (autocratic leadership)”; “Has a vision and imagination of the future (charismatic leadership)”; “Follows established rules and guidelines (self-protective leadership)”; “Concerned with the welfare of the group (team-oriented leadership)”; “Willingness to give time, money, resources and help to others (humane oriented)”; and “Acts independently, does not rely on others (autonomous leadership).” 4
Analyses and Results
In the GLOBE project, individual respondents were nested within organizations and the organizations were nested in societies. Hence, to properly analyze the GLOBE data in the present study, we conducted random coefficient models. As discussed by van de Vijver and Leung (1997) and Bliese and Hanges (2004), random coefficient modeling (RCM; also frequently referred to as HLM) accounts for the correlated errors in observations created by having nested data structure. If the nested structure of the data is ignored when conducting statistical analyses, the standard errors in these statistical analyses would be inappropriate (Bryk & Raudenbush, 1992) and result in inflated Type 1 or Type II errors, depending on the level of analysis of the hypothesis being examined (Bliese & Hanges, 2004). Hierarchical Linear Modelling (HLM) takes the nested structure into account and thus properly adjusts the standard errors for the nested data structure (Cheslock & Rios-Aguilar, 2011).
Given the nested structure of the GLOBE data, we conducted a three-level RCM to test our hypotheses. The first level was at the individual level and provided individual ratings for the CLT dimensions. The second level was at the organizational level and provided the organizational practices measures. The final and highest level was at the societal level and contained societal level scores of societal tightness and the GLOBE culture measures. We used grand mean centering in our RCM because this centering technique controls for the contribution of any lower (i.e., organizational) level predictors test before testing the unique contribution of any upper level (i.e., societal culture or tightness) predictors (Hofmann, 1997). All of our hypotheses were tested using random intercept/fixed slope RCMs.
To assess the need for analyzing the data with RCM, we conducted separate null models (i.e., no predictors in equation) for each global CLT dependent variable. The results from these models enable the estimation of the percentage of variance (ICC1) of each dependent variable operating at the organizational and societal levels of analysis. RCM is increasingly important as the magnitude of ICC1 departs from 0. LeBreton and Senter (2008) suggested that ICC1 values of .01, .10, and .25 represent small, medium, and large values, respectively. The ICC1s for our global CLTs support the use of RCM, more specifically, participative CLT, ICC1Organization = .07, ICC1Society = .16; autonomous CLT, ICC1Organization = .02, ICC1Society = .16; charisma CLT, ICC1Organization = .06, ICC1Society = .15; self-protective CLT, ICC1Organization = .08, ICC1Society = .17; team-oriented CLT, ICC1Organization = .07, ICC1Society = .15; humane-oriented CLT, ICC1Organization = .04, ICC1Society = .10. Given that we are testing the relationships between societal tightness–looseness and the six global CLTs using a random intercept/fixed slope RCM, it is important that more variance of the global CLTs also operates at the societal level than at the organizational level of analysis.
We conducted our RCM via a two-step process. In the first step, we simultaneously entered in-group collectivism, power distance, and future orientation cultural practice measures at both the organizational and societal levels. The second step of our analysis consisted of adding cultural tightness to the organizational and societal cultural practice variables. This enabled a test of the unique contribution of cultural tightness. Each hypothesis is supported when the cultural tightness coefficient in Step 2 is significantly different from zero.
In addition to examining the significance of the cultural tightness coefficient, we also computed a chi-square test of differential fit between Steps 1 and 2. This chi-square test has 1 degree of freedom and a significant chi-square value indicates that the Step 2 model fits the data significantly better than the Step 1 model. Thus, this provides another demonstration that cultural tightness adds unique explanatory power to the prediction of the CLT dimensions.
Table 2 presents the results. The first hypothesis predicted that cultural tightness would be negatively related to participative leadership CLTs. As can be seen in this table, neither the cultural tightness coefficient nor the chi-square differential fit test was significant. Thus, this hypothesis was not supported.
Unstandardized Random Coefficients Predicting the Global CLT Dimensions Using Robust Standard Errors.
Note. dfOrganization = 396, dfSociety = 25; Step 2 dfOrganization = 396, dfSociety = 24. CLT = culturally endorsed leadership theory.
p ≤ .10. *p ≤ .05. **p ≤ .01; Step 1.
Hypothesis 2 predicted that cultural tightness would be positively related to autonomous leadership. As seen in Table 2, there was evidence of a unique positive relationship between tightness and perceptions of the effectiveness of autonomous leadership relationship. Both the cultural tightness coefficient and the chi-square differential fit test were significant, supporting Hypothesis 2. People from culturally tight societies believe that confident leaders who do not seek collaboration from colleagues are more effective than people from loose cultures.
Hypothesis 3 predicted that cultural tightness would be negatively related to charismatic CLTs. As seen in Table 2, both the cultural tightness coefficient and the chi-square differential fit test were significant. There was a unique negative relationship between cultural tightness and charismatic leadership after controlling for organizational and societal cultural practices. Thus, Hypothesis 3 was supported.
Hypothesis 4 predicted that there would be a unique positive relationship between societal tightness and self-protective leadership. As shown in Table 2, neither the cultural tightness coefficient nor the chi-square differential fit test was significant. Thus, this hypothesis was not supported.
The final two columns in Table 2 show the exploratory results for the two remaining GLOBE CLT dimensions. With regard to team orientation, both the cultural tightness coefficient and the chi-square differential fit test were significant. Thus, there is evidence that people from culturally tight societies view leaders who are team oriented as less effective than do loose cultures. Finally, neither the cultural tightness coefficient nor the chi-square differential fit test was significant for the humane-oriented CLT dimension. 5
Discussion
There has been extensive research investigating the influences of cultural values on the desired attributes (i.e., mental models) of effective leaders (Aktaş & Sargut, 2011; Dickson et al., 2003; Dorfman & House, 2004; House et al., 1997; Sargut & Aktaş, 2012). This research expands on this tradition to examine the influence of cultural tightness–looseness on perceptions of effective leaders. Building on theory and research on culture and implicit leadership theory, we proposed that cultural tightness–looseness would have a unique effect on perceptions of effective leaders above and beyond other cultural dimensions.
The results indeed show that tightness is positively related to perceptions that autonomous leadership is effective and negatively related to perceptions that team-oriented leadership is effective. We theorized that individuals in tight cultures prefer autonomous leaders who do not rely on others, perhaps because they view them as strong leaders who make their own “quick” decisions, which is valued in cultures that are characterize by a high need for closure. Our data also show a negative relationship between tightness and the perceived effectiveness of charismatic leadership. This supports the notion that the status quo and prevention orientation of the tight cultures (Gelfand et al., 2006) renders visionary and inspirational leadership attributes to be seen as less effective as compared with loose cultures that focus more on innovation (Chua et al., 2015). Notably, the relationship between tightness–looseness and perceptions of effective leadership was found even after controlling for in-group collectivism, power distance, and future orientation at the societal and organizational level, illustrating that tightness–looseness has a unique influence on perceptions of effective leadership. Our findings also demonstrate that tightness–looseness has different relationships with leadership as compared with other cultural dimensions. For example, whereas in-group collectivism is positively related to the effectiveness of team-oriented leadership, tightness is negatively related to this leadership dimension.
The support for these hypotheses is notable given that we combined data from two independent studies. Gelfand et al. (2011) collected tightness–looseness scores from university students as well as individuals from a wide range of occupations during 2002 to 2004, whereas the GLOBE CLTs were gathered from middle managers who worked in organizations from three types of industries during 1997 to 2001. It is reasonable to assume that differences in population characteristics as well as differences in the time when data collection occurred added noise, resulting in an underestimate of the predicted relationships. Thus, the magnitudes of the obtained results are probably conservative estimates.
In contrast to our predictions, we did not find support for any relationship between tightness–looseness and participative leadership or self-protective leadership. This is fairly surprising given that the ICC1 results showed that both of these CLT variables varied substantially at the societal level. We note that the zero-order correlations illustrate that tightness–looseness is significantly correlated with participation, r(29) = −0.41, p < .01, and self-protective, r(29) = 0.38, p < .01, CLTs. However, once we controlled for other dimensions of societal and organizational culture (most notably societal in-group collectivism, see Table 2), the previous correlations were no longer significant. Thus, it appears that societal collectivism is perhaps a more immediate antecedent for the importance of participative and self-protective CLT than societal tightness–looseness. Future research should seek to replicate these results with other samples and methods.
In addition to making a theoretical contribution, this research also has implications for leadership practice. Autonomous leadership styles are perceived as more effective in tight (compared with loose) cultures, whereas charismatic and team leadership attributes are perceived as more effective in loose (compared with tight) cultures. This suggests that to be seen as effective, leaders in tight cultures should focus on being a strong confident independent leader who emphasizes stability, whereas leaders in loose cultures should emphasize empowerment and change as well as team orientation. It also suggests that leaders who cross between tight and loose cultures may face difficulty if they attempt to enact leadership practices that were effective in their home culture. Accordingly, global leadership training programs would benefit from including a focus on tightness–looseness in addition to other value dimensions.
This research also opens up a number of interesting directions. Future research would benefit from examining the implications of tightness–looseness for other leadership frameworks beyond implicit theories of leadership, such as substitutes for leadership, shared leadership, and transformational leadership. It would also be interesting to examine tightness and punishment of leaders. For example, would leaders’ wrongdoing be judged more harshly in tight cultures where norms are strong, or would followers allow leaders more latitude in such contexts given that leaders who are autonomous and “do their own thing” are perceived to be more effective? From a change perspective, it would also be interesting to examine the processes through which people challenge leaders and ultimately cause changes in leadership. Research has shown that people in loose cultures expect that when change occurs, it happens incrementally, whereas people in tight cultures expect that when change occurs, it has to be catastrophic (Gelfand et al., 2011). This suggests that challenges to leadership might be non-linear in tight versus loose cultures. It is also interesting to examine pendulum shifts in leadership in tight cultures. For example, in contexts that have been historically tight—where autonomous leaders are seen as effective and charismatic leaders as less effective—it would presumably take time for leaders who espouse new visions and change, and a focus on team-oriented leadership to be accepted. Finally, though this research examined a particular leadership context—that of business leaders—it would be fruitful to examine whether the effects found would generalize to other contexts. For example, if we had specified the leader as a political leader, a religious leader, or a leader in some other domains, the relationship between tightness–looseness and perceived leader effectiveness might change, or the relationship between tightness and effective leadership might be stronger in some contexts as compared with others.
As with all research, this research has strengths and weaknesses. Our sample size and the diversity of the countries included in this study is a strength and increases our confidence in our findings. Our model controlled for variables at multiple levels of analysis, which is also a notable strength. However, because the GLOBE research was conducted in only three industrial sectors (food processing, financial services, and telecommunications), our conclusions may have limited generalizability. There remains some room for speculation whether, for instance, data obtained from other sectors might have shown different results. It is important to note, however, that the GLOBE project has found rather strong correlations between questionnaire data and unobtrusive measures of cultural dimensions from other domains of life (Den Hartog et al., 1999). Given that unobtrusive measures are country specific, not sector specific, this would suggest that our results reflect national differences, rather than industry differences (Koopman, Den Hartog, & Konrad, 1999).
In conclusion, previous research on culture and leadership has focused almost exclusively on cultural values. This research expands our conceptual toolkit and broadens the scope of cross-cultural leadership research by illustrating that cultural tightness is a predictor of perceptions of effective leadership. It opens up a number of interesting new directions for leadership research in tight and loose cultures and practical implications for leadership crossing the tight–loose divide.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was based on work supported in part by the U.S. Army Research Laboratory and the U.S. Army Research Office Grant W911NF-08-1-0144.